Receptor mediated gene delivery using targeted liposomes and Quantum Dots
Dissertation
zur Erlangung des Doktorgrades
der Mathematisch-Naturwissenschaftlichen Fakultäten der Georg-August-Universität zu Göttingen
vorgelegt von Valeria Sigot aus Rosario, Argentinien
Göttingen, 2008
Korreferent: Prof. Dr. Rüdiger Hardeland Tag der mündlichen Prüfung: 02.07.08
Summary
The design of liposomal carriers for gene or drug delivery is an active field in current biomedical research. An important issue is the targeting of the carriers to specific cell types for which the therapeutical approach is intended.
The combination of liposomes and imaging probes such as Quantum Dots (QDs) in a single nanoparticle offers an excellent tool for increasing the understanding of the mechanisms of liposome uptake thereby increasing the efficiency of gene or drug delivery.
In the research developed in this thesis, biotinylated lipid particles (BLP) were loaded with plasmid DNA and labeled on the surface with QDs functionalized with the ligand EGF (Epidermal Growth Factor). In a second and novel approach, BLP were loaded with a red emittingQD655 and surface coated with a green emitting QD525 tagged with the EGF ligand.
This latter approach was intended to monitor the specificity of BLP binding and uptake in live cells by fluorescence confocal microscopy. Both types of BLP were targeted to A431 cells, a cancer cell line that overexpresses the EGF receptor (EGFR). Furthermore, BLP formulations differing in the content of the protective polymer PEG (polyethylene glycol) were tested in order to optimize the intracellular delivery of plasmid DNA or QDs.
The colocalization of green emitting QD525 and redemittingQD655, allowed the immediate discrimination of the targeted BLP from the untargeted counterparts, as well as the presence of free EGF-QD525 complexes during live cell imaging. EGFR-targeted BLP showed increased binding and accelerated uptake as compared to EGF-free BLP. Cellular binding of targeted BLP was further associated with the perinuclear localization of QDs in clusters which persisted over days, indicating the difficulty in releasing the encapsulated nanoparticles from endo- lysosomal vesicles. On the other hand, targeted BLP with encapsulated plasmid carrying the reporter gene for the Green Fluorescent Protein (GFP) showed enhanced transfection of A431 cells compared to that exhibited by non-targeted particles for all lipid formulations tested, indicating the occurrence of plasmid release in the cytoplasm. No targeting effect, however, was evident in cells devoid of EGF receptor or in the receptor occupied by competing free ligand, indicating that enhanced EGFR-targeted delivery is primarily mediated via a ligand/receptor interaction.
From a therapeutic point of view, the specificity displayed by the BLP labeled with two colors of QDs and targeted to tumor cells overexpressing the EGF receptor may provide a platform for testing the specific delivery of tumorigenic drugs. This approach was also intended to elucidate the fate of lipid particles in real time, taking advantages of the photostability and bright fluorescence of QDs.
The designed BLP will provide refined information about the still poorly understood trafficking processes and the subcellular barriers to gene or drug delivery via liposomal carriers.
Table of contents
Abbreviations ... 1
1. Introduction and aims of the thesis ... 3
1.1. Molecular therapy ... 5
1.1.1. Gene delivery ... 5
1.2. Viral and non-viral gene delivery systems ... 6
1.2.1. Viral vectors... 6
1.2.2. Non-viral vectors... 7
1.3. Lipid-based delivery systems... 8
1.3.1. Components of lipid-based nanoparticles ... 8
1.3.2. Sterically Stabilized Liposomes ... 9
1.3.3. Stabilized Plasmid-Lipid Particles (SPLP) ... 10
1.4. Cell mechanisms of liposome uptake ... 11
1.4.1. Receptor-mediated endocytosis as a route for targeted liposomal delivery... 13
1.4.2. EGF receptor targeted delivery systems... 14
1.5. Fluorescent nanoparticles in live-cell imaging ... 14
1.5.1. Applications for lipid-based delivery systems ... 14
1.5.2. Long-term imaging using Quantum Dots ... 15
1.5.3. Ligand tagged Quantum Dots... 16
1.6. Aims and organization of the thesis ... 18
2. Materials and methods... 21
2.1. Reagents ... 23
2.1.1. Lipids ... 23
2.1.2. Biotin and biotin-related reagents... 24
2.1.3. Buffers ... 24
2.1.4. Quantum Dots... 24
2.1.5. Labeled proteins and organic fluorophores... 24
2.1.6. Plasmid DNA... 25
2.1.7. bis-PNA (bis Peptide Nucleic Acid) ... 25
2.1.8. Restriction endonucleases and DNA standards... 25
2.2. Cell lines, media and labware... 25
2.3. Miscellaneous ... 25
2.4. Equipments and softwares ... 26
2.5. Preparation of Biotinylated Lipid Particles (BLP) ... 27
2.5.1. Detergent dialysis technique...27
2.5.2. Ultracentrifugation in sucrose density gradient... 28
2.5.3. Particle size analysis by Dynamic Light Scattering... 28
2.5.4. DNA quantitation ... 29
2.5.5. Biotin quantitation ... 30
2.5.6. Transmission Electron Microscopy... 31
2.6. Labeling and targeting of BLP ... 32
2.6.1. EGF-QDs preformed complexes ... 32
2.6.2. EGF-QDs coupling to BLP ... 32
2.7. Live-cell experiments ... 33
2.7.1. Cell culture ... 33
2.7.2. Experimental conditions for the incubation of A431 cells with BLP ... 33
2.7.3. Incubation of EGF-QDs and Transferrin-Alexa488 complexes with A431 cells... 33
2.7.4. Competitive binding assay... 34
2.7.5. Experimental conditions for targeted gene delivery by BLP ... 34
2.8. Confocal fluorescence microscopy ... 35
2.8.1. Simultaneous detection of two different colors of QDs ... 35
2.8.2. Colocalization analysis ... 35
2.9. Quantitative GFP expression analysis ... 36
2.9.1. Epi-fluorescence microscopy... 36
2.9.2. Flow cytometry ... 37
2.10. bis-PNA:DNA Hybrids ... 38
2.10.1. Plasmid DNA amplification and purification... 38
2.10.2. Biotinylation of bis-PNA... 38
2.10.3. Cy5 labeling of bis-PNA... 38
2.10.4. Hybridization of bis-PNA to pEGFP-C1 plasmid... 38
2.10.5. QD labeling of Hybrids... 39
2.10.6. Restriction fragment analysis by gel electrophoresis... 39
2.10.7. Conjugation of QD-Hybrids to biotin coated magnetic beads ... 40
2.11. Atomic Force Microscopy analysis ... 40
2.11.1. Sample preparation for scanning... 41
3. Results: “Characterization of Biotinylated Lipid Particles”... 43
3.1. Encapsulation of plasmid DNA and QDs in BLP ... 45
3.1.1. BLP formulations... 45
3.1.2. Time course formation of BLP during detergent dialysis... 46
3.1.3. BLP-DNA purification from non-encapsulated DNA ... 47
3.1.4. Effect of PEG content in DNA encapsulation efficiency and particle size ... 49
3.1.5. BLP size analysis by AFM ... 50
3.1.6. Encapsulation of QDs in BLP by the detergent dialysis method ... 51
3.1.7. Characterization of fluorescent BLP-QDs ... 52
4. Results: “Cell uptake of EGFR targeted and QDs-labeled BLP” ... 55
4.1. Targeting and labeling of BLP ... 57
4.1.1. Cell uptake of targeted BLP-QDs... 58
4.1.2. Intracellular distribution of internalized BLP ... 59
4.1.3. Time course of BLP internalization... 60
4.1.4. EGFR targeted BLP follow the route of EGF-QDs complexes ... 62
4.1.5. Intracellular fate of QDs encapsulated in targeted BLP... 64
4.2. Receptor specificity of targeted BLP ... 65
4.2.1. Competitive assay ... 65
4.2.2. Interaction of BLP with EGFR-negative cells... 66
4.3. Delivery and transfection abilities of targeted and QDs labeled BLP-DNA ... 67
4.3.1. Intracellular fate of dual-labeled and targeted BLP-DNA... 67
4.3.2. Transfection competency of dual-labeled and targeted BLP-DNA... 69
4.3.3. BLP targeting in a two-step procedure... 70
5. Results: ”Single Quantum Dot labeling of plasmid DNA” ... 73
5.1. bis-PNA strand invasion of supercoiled plasmids ... 75
5.2. Quantum Dot labeling of biotin-bis-PNA: plasmid DNA Hybrids ... 77
5.2.1. Specificity of single QD labeling revealed by AFM... 78
5.2.2. Visualization of Hybrids and QD:Hybrids complexes on magnetic beads... 79
6. Discussion ... 83
6.1. PEGylated and Biotinylated Lipid Particles ... 86
6.2. Single step conjugation for targeting and labeling of BLP ... 87
6.3. Non-covalent Quantum Dot labeling of plasmid DNA... 90
6.4. Rational liposome design... 91
7. Conclusions ... 93
References... 97
Acknowledgments... 105
Publications related to this thesis project ...106
Abbreviations
AFM Atomic Force Microscopy
a.u. arbitrary units
BLP Biotinylated Lipid Particles
bp base pair
BP Band pass filter (optics)
BSA Bovine serum albumin
CCD Charged-coupled device
CHO Chinese hamster ovary cell line
CLSM Confocal Laser Scanning Microscopy
cmc critical micellar concentration
DIC Differential interference contrast
DLS Dynamic Light Scattering
DMEM Dulbecco modified Eagle medium
DNA Deoxyribonucleic acid
DOPE Dioleoyl-L-a-phosphatidylethanolamine DOTAP 1,2-Dioleoyloxy-3-(trimethylammonio)-propane DSPE Distearoyl-L-a-phosphatidylethanolamine
EDTA Ethylenediamine tetraacetic acid
EGF Epidermal growth factor
EGFR Epidermal growth factor receptor
ex./em. excitation/emission wavelengths
FCS Fetal calf serum
g gravitational field centrifugation unit
GFP Green fluorescent protein
HBS HEPES buffered saline
HEPES N-(2-hydroxyethyl)piperazine-N´-(2-ethanesulfonic acid)
kDa kilodalton
LB Luria Bertani growth medium
LP Long pass filter (optics)
mM, μM, nM millimolar, micromolar, nanomolar
min, h minute, hour
MWCO Molecular Weight Cut-off
NA Numerical aperture
OGP Octylglucopyranoside
PBS Phosphate buffered saline
pdi Polydispersity index
PEG Polyethyleneglycol
PEG-Cer-C20 PEG-Ceramide bioconjugate with an arachidoyl acyl group PEG-Cer-C8 PEG-Ceramide bioconjugate with an octanoyl acyl group PEG-DSGS PEG-[succinoyl]-1,2-distearoyl-sn-glycerol
pH Negative decadic logarithm of the hydrogen ion (H+)
concentration
PNA Peptide nucleic acid
QD Quantum dot
RT Room temperature
SPLP Stabilized Plasmid Lipid Particles StAv Streptavidin t time
TEM Transmission Electron Microscopy
UV Ultraviolet
1. Introduction and aims of the thesis
1.1. Molecular therapy
The genomic information obtained through the Human Genome Project has accelerated enormously the analysis of the functions of various disease relevant genes as well as the design of therapeutic molecules and delivery systems (Hicks, 2003; Austin, 2004).
Biomolecules, such as sense and antisense oligonucleotides, small interference RNAs and peptides, as well as entire genes (cDNA) and proteins are essential tools in the development of molecular therapies (Shi and Hoekstra, 2004; Wagner, 2007). However, the potential of such information-rich macromolecules for therapeutic use has been limited by the difficulties in delivering them specifically to target tissues in vivo and in discriminating between disease- affected and normal cells.
A variety of viral and non-viral vehicles have been conceived to effect the encapsulation, in vivo cell targeting and intracellular delivery of therapeutic macromolecules (Maurer et al., 1999; Gupta et al., 2005; Torchilin, 2006). However, the hostile extracellular environment constitutes a primary challenge in the systemic delivery of gene or drug carriers to intracellular targets. This presents a number of barriers, such as extreme pH, proteases and nucleases and the immune defence and scavenger systems (Figure 1.1). Once the carrier reaches the target tissue, the efficient cell binding, uptake and delivery of the cargo into and within subcellular compartments is critical for gaining the desired therapeutic effect. At this stage, the cell membrane is the main physical barrier due to the poor permeability of the lipid bilayer, where the diffusion rate is highly dependent on the size and hydrophobicity of the molecular carrier. Therefore, the rational design of intracellular gene/drug delivery vehicles requires a better understanding of the cell membrane barrier (Belting et al., 2005).
1.1.1. Gene delivery
The realization that nucleic acids could be used as therapeutic agents to alter the expression and function of proteins in the body has given an immense boost to transfection biology. Gene therapy is a part of the growing field of molecular medicine gaining significance in the treatment of human diseases. By delivering genes into cells to regulate or supplement defective genetic loci, the morbidity of inherited and acquired diseases can be alleviated (Civin, 2000; Edelstein et al., 2007).
In most procedures to date, nucleic acids have been encapsulated in vehicles that both protect the therapeutic gene and allow its extracellular and intracellular trafficking. Most DNA delivery systems operate at one of three general levels: DNA condensation and complexation, endocytosis, and nuclear membrane translocation (Luo and Saltzman, 2000).
Negatively charged DNA molecules are usually condensed and/or complexed with cationic lipids or polymers before the delivery into cells. These complexes are taken up by cells usually
through an endocytic mechanism, which determines subsequent DNA release, trafficking, lifetime and function in the cell (Guy et al., 1995).
Figure 1.1: Illustration of barriers for systemic gene delivery. DNA-containing nanoparticles are injected intravenously into human body. Serum proteins may bind to the particles, crosslink them and increase the particle size. This can result in rapid particle elimination (Insert 1). The Kupffer cells (RES) may take up particles, leading to rapid nanoparticles elimination from circulation and decrease their access to the hepatocytes (Insert 2). Circulating nanoparticles may extravasate in tumor tissue through the leaky tumor vessels (Insert 3). Particles then need to pass through the crowded extracellular matrix to contact the cell surface (Insert 4). When the particles are internalized into cells, DNA must escape from the endosome and find its way into the nucleus (Insert 5). Adapted from (Weijun Li, 2007).
1.2. Viral and non-viral gene delivery systems
1.2.1. Viral vectors
Viruses in nature have evolved an exceptional ability for delivering their genome into cells, responding to changes in the cellular environment and gaining access to their desired intracellular compartments (Vives et al., 2006).
A number of different viruses, such as retrovirus, adenovirus, adeno-associated virus and herpes virus have been widely studied in gene transfer systems attracting the most attention in the field of transduction (Zhang and Godbey, 2006). Retroviral vectors were the first viral
vectors to enter clinical trials and continue to be attractive candidates for applications where integration of the transgene is required (Cornetta K, 2008).
Virus-based vectors are modified by removing certain viral genes associated with pathogenic function. Other approaches mimic the viral machinery required for cell binding and internalization by coupling modified proteins from the viral capside to the surface of therapeutic carriers (Boeckle and Wagner, 2006). In fact, more than 70% of recent clinical protocols involving gene therapy use recombinant virus-based vectors for DNA delivery (Cornetta K, 2008). However, these delivery systems are restricted by their limited DNA carrying capacity, toxicity and potential oncogenicity, factors that hamper their routine use in basic research laboratories (Anderson, 1998; Boeckle and Wagner, 2006).
1.2.2. Non-viral vectors
Synthetic non-viral vectors have many potential advantages over viral ones, including a size-independent delivery of nucleic acids ranging from oligonucleotides to artificial chromosomes, significantly reduced toxicity and immunogenicity, and lower potential for oncogenicity (Luo and Saltzman, 2000). Additionally, they involve less demanding quality controls, and less complex pharmaceutical and regulatory requirements. However, these reagents still demonstrate reduced transfection efficiencies and transient levels of gene expression compared to the viral vehicles (Lechardeur et al., 2005; Mehier-Humbert and Guy, 2005).
Non-viral approaches include physical methods for introducing naked DNA directly into the cell, such as particle bombardement, microinjection and electroporation (Chou et al., 2004; Mehier-Humbert and Guy, 2005). The drawbacks of these systems are the limited number of cells that can be treated and the possible physical damage to the cell membrane.
Widespread approaches involve chemical methods to enable the cellular uptake of insoluble or highly charged macromolecules. Chemical carriers are based on peptides, lipids and phospholipids mixtures as well as polymers (Kramer et al., 2004; Kneuer et al., 2006) which form complexes with the cargo molecules such as drugs or DNA. In these delivery systems, the cargo does not get into contact with the cell surface molecules, so that unspecific interactions are omitted. Chemical carriers enhance the uptake of macromolecules by facilitating either direct cell translocation (Merkleb, 2004) or endocytosis (Hoekstra et al., 2007). Lipid or polymer-based delivery systems have now become the major tools to introduce not only nucleic acids but also proteins, peptides and nanoparticles into cultured cells.
The scope of non-viral delivery systems have been expanded by the incorporation of environment-sensitive polymers, carefully designed to take advantage of intracellular changes in pH and temperature in order to achieve a controlled release of the therapeutic
molecule (Heath et al., 2007). Furthermore, to achieve active transport of DNA through the nuclear pore complex, nuclear localizing signal (NLS) peptides have been widely used (Branden et al., 2001; Simonson et al., 2005). In most cases, NLS is conjugated with a gene carrier such as polyethylenimine (PEI) or with DNA directly.
1.3. Lipid-based delivery systems
The most studied and widespread non-viral approaches for deliverying DNA have been those based on lipids.A breakthrough was achieved when Felgner reported for the first time that complexes of plasmid DNA and the cationic lipid dioleyl-trimethylammmonium chloride (DOTMA) at a 1:1 molar ratio with dioleoylphosphatidylethanolamine (DOPE), were avidly internalized by cells and caused profound expression of the plasmid (Felgner et al., 1987).
Cationic lipid/nucleic acid complexes or lipoplexes have been the subject of intensive investigation in recent years. The main focus has been the characterization of the molecular mechanisms of lipid-vector and cell membrane interaction, essential for overcoming intracellular barriers (Felgner and Ringold, 1989; Simoes et al., 2005; Wasungu and Hoekstra, 2006; Hoekstra et al., 2007).
Originally developed for the transfection of DNA, lipid vectors have become the major tools for introducing not only a variety of nucleic acids but also proteins, peptides and nanoparticles into cells in vitro and in vivo (Maurer et al., 1999; Audouy and Hoekstra, 2001;
Torchilin et al., 2003).
The other major field of lipid-based delivery systems arose from the observation that phospholipids in aqueous systems can form closed bilayered structures, known as liposomes (Lasic and Papahadjopoulos, 1995). These particles comprising an outer lipid layer membrane surrounding and internal aqueous space can be loaded with therapeutic biomolecules (Mayer et al., 1989). Liposomes can encapsulate and facilitate the delivery of plasmid DNA containing therapeutic genes with sizes of several kilobases (Fenske and Cullis, 2005).
Liposomes represent a flexible platform for encapsulation since they can range from multilamellar vesicles (MLVs) with diameters of several microns to small unilamellar vesicles (SUV) of about 20 nm. For biomedical applications, particles with the greatest utility have diameters of ~100 nm, large enough to carry a significant payload but small enough to ‘slip’
between leaky endothelial junctions escaping the immune system surveillance (Szoka and Papahadjopoulos, 1980).
1.3.1. Components of lipid-based nanoparticles
Usually, the nanosized lipid particles are constructed from a combination of synthetic and natural lipids, lipopolymers and pH or reduction-sensitive components. For most of these lipid-
Barenholz, 1999). An added advantage of the cationic lipids is that they can bind to negatively charged mammalian cell membranes inducing the uptake of the associated nucleic acid into cells. Neutral lipids such as the fusogenic lipid 1,2-dioleoyl-sn-glycero-3- phosphoethanolamine (DOPE) and cholesterol are usually used as helper lipids which may increase transfection activity of the DNA-containing lipid-based carrier (Hafez and Cullis, 2001). For example, DOPE has a phosphoethanolamine head group whose size is smaller than its hydrophobic diacyl chain facilitating membrane fusion and disruption, and consequently the DNA release (Zuhorn et al., 2005).
1.3.2. Sterically Stabilized Liposomes
The fast and efficient clearance of conventional liposomes from the circulation by macrophages has seriously compromised their application for the treatment of a wide range of diseases involving other organs. The incorporation of the hydrophilic and biocompatible polymer polyethylene glycol (PEG) improved the lifespan of liposomes in the bloodstream and reduced cell toxicity, leading to a renewed interest in liposomal carriers during the 90’
(Lasic et al., 1991). PEG has been covalently linked to natural and synthetic lipids, which were subsequently incorporated into liposome formulations (Bhadra et al., 2002). The PEG polymer acts like a shield protecting the encapsulated therapeutic agent from enzymatic degradation, rapid renal clearance and interactions with cell surface proteins, thereby minimizing adverse immunological effects (Woodle and Lasic, 1992). Figure 1.2 schematically depicts a conventional liposome and the advanced sterically stabilized liposome surface- coated with PEG.
Figure 1.2: Schematic representation of a conventional liposome and a sterically stabilized PEG- liposome. Adapted from Encyclopedia Britannica (online).
‘Pegylation’ is now established as the method of choice for improving the pharmacokinetics and pharmacodynamics of liposomes. However, once the liposomal carrier reaches the cell surface, pegylation may interfere for instance with the endocytic mechanism of uptake by either lowering the binding affinity to cell receptors or by preventing the intermembrane contact between liposomal surface and endosomal
membranes, required to release the cargo in the cytoplasm (Shi et al., 2002; Song et al., 2002). Therefore, PEG-lipids included in liposomal formulations should dissociate from the complex, once the liposome is in close contact with endosomes upon internalization. This has been accomplished by incorporating a short acyl chain lipid-PEG to the liposome formulation with the ability to promote within minutes the lipid transfer from the liposome bilayer (Harvie et al., 2000). Alternatively, by adding a cleavable, pH-sensitive PEG analogue in which the polymer moiety is cleaved off upon exposure to the acidic environment of certain endosomal compartments (Kirpotin et al., 1996; Guo and Szoka, 2001; Shi et al., 2002;
Choi et al., 2003; Guo et al., 2003).
Current development is focused on combining long-term circulating (Lasic et al., 1999) and targeted liposomes (Sofou and Sgouros, 2008). The later feature, achieved by incorporating a suitable ligand or antibody to the liposome surface, is essential for improving the specificity for target cells.
1.3.3. Stabilized Plasmid-Lipid Particles (SPLP)
Employing a detergent dialysis technique (Hofland et al., 1996), plasmid DNA has been encapsulated with a 50-70% efficiency in small particles (~70 nm) stabilized by PEG-Ceramide lipids (Wheeler et al., 1999). These nanocarriers, known as a Stabilized Plasmid Lipid Particles (SPLP), have been further improved by several groups by adding pH-sensitive lipids and ceramides with shorter carbon chains to confer endosamal escape capabilities (Mok et al., 1999; Wheeler et al., 1999; Fenske et al., 2002; Li et al., 2005).
The first generation SPLP developed by Wheeler et al. contained a high percentage of the helper lipid DOPE (84 mol%), low levels of cationic lipid dioleyl-dimethylammonium chloride DODAC (6 mol%) and quite high levels (10 mol%) of PEG-Ceramide with an arachidoyl acyl group (PEG-Cer-C20). The release half-life (t1/2) of PEG-Cer-C20 from SPLP was 13 days, becoming an intractable steric barrier to transfection. The length of the ceramide lipid anchor determined the time that the PEG conjugate remained associated with the bilayer. When PEG-Cer-C20 was replaced by PEG-Cer-C8 with an octanoyl acyl group, the t1/2 was reduced to 1.2 min, considerably increasing the transfection of cells in vitro (Mok et al., 1999). This first generation of SPLP was improved by Szoka and coworkers (Choi et al., 2003; Li et al., 2005) by adding a pH sensitive PEG-lipid to trigger plasmid release in the acidic endosomal environment. The key design feature of such PEG-lipids is that the pH triggerable PEG-linker should be completely stable at pH 7 and sufficiently destabilized at pH 5.5, as to irreversibly dissociate within at least 1 hour, essential requirement to ensure quantitative release of nucleic acid from the lipid particle.
As mentioned above, entrapment of plasmid DNA in SPLP was accomplished by a detergent dialysis procedure (Hofland et al., 1996) by which reconstituted liposomes consist
detergent is being removed, a series of micelle-micelle interactions are initiated to minimize the unfavourable energy resulting from the exposure of lipids to the aqueous medium (Figure 1.3 I). At a critical micelle size, the amplitude of the bending is sufficient to cause bilayer closure (Figure 1.3 II) and plasmid encapsulation (Figure 1.3 III). One of the most important aspects of SPLP is their great structural integrity with negligible variation in size or DNA encapsulation after 5 months at 4 °C.
Figure 1.3: Schematic representation of SPLP formation during detergent dialysis. Detergent removal from lipid-detergent micelles causes the transformation of small micelles into larger ones (I), which bend upon further detergent removal to form curved mixed micelles and trapp the plasmid DNA (II). The bilayer becomes continuous forming the stable plasmid-lipid particles (III). Adapted from (Jean-Louis Rigaud, 1998).
SPLP have been designed for long-term circulation in vivo but given the severity of the extracellular environment, only a modest detectable transfection in animals post i.v.- administration has been reported (Monck et al., 2000). SPLP systems are now under evaluation in Phase I clinical trials (Protiva, unpublished data).
1.4. Cell mechanisms of liposome uptake
The cellular uptake of liposomes is generally believed to be mediated by adsorption of liposomes onto the cell surface and subsequent endocytosis. As previously described, the
rate limiting steps in this process are the efficiency of cell surface association, internalization, release of loaded drug/genes from intracellular compartments such as endosomes, transfer to the cytosol and eventually, translocation into the nucleus.
Liposomal carriers can adsorb specifically and non-specifically to the cell surface (Figure 1.4 a, b), in some cases promoting a direct delivery of the therapeutic agent into the cytoplasm (Figure 1.4 c, d). Additionally, in order to increase the initial cell binding, delivery carriers have been modified with cationic moieties to increase the electrostatic interaction of the vehicle with the negatively charged cell membrane (Figure 1.4 e, c). Constructs that are more sophisticated involve the covalent binding of a ligand or antibody to the liposome carrier to promote the specific receptor-mediated or antigen-mediated endocytosis, respectively (Figure 1.4 a). If uptake occurs via endocytosis (Figure 1.4 f), receptor or non- receptor mediated, the liposomes end up in endosomal compartments.
In the case of receptor-mediated endocytosis, internalized vesicles and their content are sorted mainly to the lysosomal compartment for degradation of the ligands and the loaded cargo (Figure 1.4 f, g). This also accomplishes the downregulation of activated receptors (Shepherd, 1989). However, the majority of targeted deliveries aim to avoid lysosomal trafficking in an effort to protect the drug molecule or biomolecules from enzymatic degradation.
Lysosomes, as well as the endoplasmatic reticulum (ER), Golgi apparatus, and endosomes are, to a variable degree topologically continuous with the exterior of the cell, i.e. the exit of macromolecules from any of these compartments requires passage through a lipid bilayer.
In general, the endocytosis mechanisms can be divided into clathrin-dependent and clathrin-independent. There is a close relation between particle size and the preferred internalization pathway, although the correlation is not absolute. Rejman et al. proposed that particles <200 nm are taken up via the clathrin-dependent mechanism whereas larger particles (250-500 nm) enter the cell through caveolin 1-rich vesicles (Rejman et al., 2004). This correlation, between particle size and uptake pathway, is indeed plausible since some of these pathways are constrained to the size of their vesicles – e. g. clathrin-coated and caveolae-derived vesicles (Pelkmans et al., 2004).
Drug or DNA
Figure 1.4: Liposome uptake and intracellular fate: DNA or drug loaded liposomes can specifically (a) or nonspecifically (b) adsorb onto the cell surface. Liposomes can also fuse with the cell membrane (c), and release their content into the cytoplasm, or can be destabilized by certain cell membrane components when adsorbed on the surface (d) so that the released drug can enter the cell via micropinocytosis. Liposomes can undergo the direct transfer or protein-mediated exchange of lipid components with the cell membrane (e) or be subjected to a specific or nonspecific endocytosis (f). In the case of endocytosis, liposomes can be directed into lysosomes (g) or, in route to the lysosome, they can provoke endosome destabilization (h), which results in the liberation of the therapeutic agent into the cell cytoplasm (Adapted from (Torchilin, 2005).
Clathrin-mediated endocytosis serves as the main mechanism of internalization for macromolecules and plasma membrane constituents for most cell types and it is the most investigated vesicular pathway for targeted drug delivery. This is initiated by the formation of clathrin-coated pits, which are subsequently pinched off and internalized. The clathrin coat is removed and multiple vesicles fuse originating the early and late endosomes that ultimately fuse with lysosomes (Mousavi et al., 2004).
1.4.1. Receptor-mediated endocytosis as a route for targeted liposomal delivery
Targeting is usually achieved by conjugating a high affinity ligand to the carrier that provides preferential accumulation of the latter for instance, in a tumor-bearing organ, in the tumor itself, in individual cancer cells or intracellular organelles. In most cases the targeting moieties (ligands or antibodies) are directed toward specific receptors or antigens exposed
on the plasma membrane (Kato and Sugiyama, 1997). The overexpression of receptors or antigens in many human cancers lends itself to efficient drug uptake via receptor-mediated endocytosis. Folate and Transferrin are widely applied ligands for liposome targeting because their cognate receptors are frequently overexpressed in a range of tumour cells (Kakudo et al., 2004; Hilgenbrink and Low, 2005). Liposomes tagged with various monoclonal antibodies have been delivered to many targets (Park et al., 2001).
The performance of non-viral vector could be certainly optimized by targeting them into distinct cellular internalization pathways, considering that not every pathway may be equally effective in releasing a therapeutic biomolecule in the cytosol. This step is critical for nucleic acid delivery to increase the possibility of nuclear transport and the ultimate expression of the delivered genes (Bareford and Swaan, 2007).
1.4.2. EGF receptor targeted delivery systems
The epidermal growth factor receptor (EGFR, ErbB) is a 170 kDa protein that is distributed randomly on the surface of cells, excluding hematopoietic cells. Its ligand, EGF, is a 53 aminoacid-peptide that mediates cellular signal events regulating cell proliferation, differentiation, cell cycle progression, adhesion, invasion, angiogenesis and inhibition of apoptosis. Following binding of the EGF, the EGFR functions either as a homodimer through the complexation of two identical EGFR molecules or as an heterodimer by associating with one of the three other ErbB family members; in either case the resulting dimerization then initiates the cellular internalization (Bublil and Yarden, 2007).
The EGFR is a tempting target for gene delivery since it is overexpressed in a wide variety of human tumors found in cancers of head and neck, breast, colon, ovary, lung, prostate and liver (Johnston et al., 2006). Enhanced EGFR expression is associated with tumor invasiveness, resistance to chemotherapy and radiation therapy and clinically correlates with poor prognosis and lower patient survival (Rubin Grandis and V. A.; Wagener, 1998).
Squamous cell carcinoma of the head and neck is a cancer commonly associated with EGFR overexpression (>90%), which appears to play a role in the unregulated growth of these cells (Cohen, 2006). EGFR has been used to target drugs and toxins loaded in therapeutic liposomes (Kullberg et al., 2003; Mamot et al., 2005) and lipoplexes (Shir et al., 2006).
1.5. Fluorescent nanoparticles in live-cell imaging
1.5.1. Applications for lipid-based delivery systems
The use of multifunctional nanoprobes for molecular and cellular imaging is already showing great promise, providing new insights into approaches such as gene therapy and drug delivery.
Underlying the rational design of gene or drug loaded delivery vectors is the recognition of the individual steps of a particular internalization pathway. Live-cell imaging and single- particle tracking using confocal fluorescence microscopy have shed light on the intracellular dynamics and bottlenecks of the drug/gene delivery process (Payne, 2007; Zhang et al., 2007).
1.5.2. Long-term imaging using Quantum Dots
Continuous cell imaging has been dramatically improved with the introduction in the last decade of fluorescent probes such as Quantum Dots (QDs), colloidal nanocrystals with unique optical properties that make them outstanding fluorescent labels for long-term and multicolor imaging (Alivisatos et al., 2005; Michalet et al., 2005).
QDs commonly consist of a cadmiumselenide (CdSe) or cadmiumtelluride (CdTe) core enclosed within a zincsulfide (ZnS) passivation shell. The cadmium-based core is toxic to cells, but the ZnS shell isolates it from the cell rendering the core non-toxic at functionally useful concentrations. Peptides, polyethylene glycol (PEG) or other biocompatible polymers, also provide protection and specific sites for bioconjugation (Figure 1.5, left top panel).
The main advantages of QDs over organic fluorophores are the greater photostability and the excitation wavelengths range that extends above 500 nm. This latter feature reduces cell phototoxicity, essential for long-term fluorescence imaging. Because the molar extinction coefficients (0.5–2 × 106 M–1cm–1) of QDs are about 10–50 times larger than those of organic dyes, the QDs absorption rates will be 10–50 times greater than those of organic dyes at the same excitation photon flux. Because of the corresponding increased rate of light emission, single QDs appear 10–20 times brighter than organic dyes (Li et al., 2007). Furthermore, the emission wavelengths of QDs are size-tunable (Figure 1.5, left bottom panel). For example, QDs of approximately 2 nm in the core diameter produce a blue emission, whereas QDs of approximately 7 nm in diameter emit red light. Therefore, multiple colors QDs can be generated by controlling the size of the nanoparticle during the synthesis. Although QDs absorption spectra are broad, emission spectra are narrow (Figure 1.5, right panel) without the extension to the red characteristic of organic dyes. This feature allows for simultaneous detection of multiple color QDs upon illumination with single light source (Smith et al., 2004).
Despite these features, QDs biocompatibility remains a critical issue to use in humans and a possible limitation to their in vivo applications (C. Kirchner and H.E. Gaub, 2005).
Figure 1.5: Scheme of Quantum Dots nanoparticles. Left top panel: QDs of CdSe core are surrounded by a ZnS shell and coated with biocompatible polymers. Left bottom panel: Size-tunable fluorescence emission. Right panel: Absorption (upper curves) and emission (lower curves) spectra of four CdSe/ZnS QDs samples. The blue vertical line indicates the 488-nm line of an argon-ion laser, which can be used to efficiently excite all four types of QDs simultaneously. Adapted from (Daniele Gerion, 2001).
1.5.3. Ligand tagged Quantum Dots
QDs have to be functionalized to act as useful, specific fluorescent labels. The targeting moiety attached to the QD surface determines the mode of entry into cells and their intracellular localization. Selective staining of plasma membrane and intracellular organelles was achieved using QD-labeled antibodies, receptor ligands or targeting peptides (Figure 1.5, left top)(Michalet et al., 2005; Al-Jamal and Kostarelos, 2007). Dynamic processes such as diffusion of membrane receptors (Dahan et al., 2003) or antigen uptake by cells (Cambi et al., 2007) have been tracked with ligand and antigen tagged QDs, respectively.
The cell binding of EGF tagged QDs has been extensively investigated in the laboratory in which this thesis work was carried out. It has been demonstrated that complexes of streptavidin-conjugated quantum dots (QDs) with biotinylated EGF are biochemically competent ligands for erbB1, the EGF receptor (Lidke et al., 2004). The application of multicolor EGF-QDs allowed tracking endocytosis of the receptor-ligand complex to follow subsequent steps in time. Figure 1.6 (A,B) shows two different colors of ligand-QDs delivered sequentially to identify early and late endosomes (Lidke et al., 2007). In addition, a previously
detected at the single molecule level (Lidke et al., 2005). QDs allowed the visualization of the displacement of EGF-EGFR complexes on filopodia towards the surface and the interior of living A431 cells (Figure 1.6 (C,D).
Figure 1.6: Tracking of endosomes with two different color QD-EGF complexes. A: QD605-EGF added to A431 cells, incubated RT for 3 min, washed and incubated for 5 min at 37° C, then QD525-EGF were added and images acquired for 8 min at 37° C. Early endosomes: red only whereas QD525-EGF is predominantly external. B: QD525-EGF washed and observed 20 min later at 37° C. Several classes of endosomes containing red only, green only or both QD colors (see inset). Early, sorting and late endosomes are loaded (Lidke et al., 2007). Retrograde transport visualized with single EGF tagged QD.
C: Selected frames of an A431-erbB1-GFP cell (green) from a time series after binding 5 pM EGF-QD (red) followed by addition of free EGF (50 ng/ml) at 300 s. Wide field image, scale bar is 5 μm. Images arecontrast enhanced. D: Trajectory of the indicated monomer erbB1-EGF-QD complex (yellow arrow) on a filopodium that exhibits random diffusional movement (black) until the addition of unlabeled EGF (green box), after which the complex exhibits active retrograde transport (red). Adapted from (Lidke et al., 2005).
The recent advances in the chemistry of QD coating permit the simultaneous conjugation of two or more different molecules, and thus, the simultaneous stimulation of several receptors, further broadening the potential applications of these fluorescent probes for dynamic studies of cellular processes (Xiaohu Gao, 2004). However, the existence of multiple groups on the surface of QDs can be considered a disadvantage in comparison to single attachment sites or reactive entities. In fact, a high number of molecules contributing with multiples sites of conjugation leads to a distribution of QDs subpopulations differing in
particle-label stoichiometries and undesirable cross-linking. Procedures for achieving precise control of the number of sites have been developed (Sperling, 2006; Lidke et al., 2007).
Multiple sites can be advantageous for multifunctionality, for instance, the combination of targeted QDs and lipid carriers in a single nanoparticle can serve to elucidate the internalization route, fate of carriers and processing of therapeutic drugs. The attractiveness of such hybrid nanoparticles are:
Lipid carrier and QDs can be surface functionalized with a ligand for specific targeting
Complexes of drug-QDs can be encapsulated in the lipid carriers improving QDs biocompatibility and providing a therapeutic effect
The fluorescent lipid vehicle can be tracked continuously
Combining two colors QDs on the surface and inside the lipid vesicle, the fate of the carrier and the loaded drug could be followed independently
1.6. Aims and organization of the thesis
A new emerging field in targeted delivery is the multi-modality carrier, in which a therapeutic molecule, an imaging agent or a fluorescent probe and a targeting entity are assembled in a single nanometer-scaled scaffold for simultaneous imaging and therapy.
The aim of this thesis was to develop a QD labeled and EGFR targeted delivery system mediated by Biotinylated Lipid Particles (BLP).
The human epidermoid carcinoma cell line A431 was chosen because of its large overexpression of EGFR, approximately 106 receptors per cell.
The original approach proposed here was not only the specific targeting but the novel encapsulation of QDs in these lipid particles by a detergent dialysis technique. A dual-color labeling strategy with loaded and surface bound QDs was introduced to follow the intracellular localization of the cargo and carrier by confocal microscopy.
EGFR targeted and QD labeled BLP were loaded with plasmid DNA and cell specificity and transfection efficiency of the delivery system were evaluated according to the expression of the GFP reporter gene.
Finally, a DNA labeling strategy was conceived to achieve sequence specific and non- covalent labeling of DNA using a single QD and bis-peptide nucleic acids (bis-PNA). The objective was to obtain a transfection competent and labeled DNA cargo for the developed BLP.
The specific objectives were:
Create a functional ligand-QDs labeled BLP
Demonstrate enhanced receptor-mediated uptake of the complex and gene expression
Develop the imaging methodology to follow intracellular localization of lipid particles and cargo by confocal microscopy
Design a QD-PNA-DNA complex for cellular imaging of DNA in delivery systems
The results obtained in this study are divided in three sections. The first one describes the preparation of biotinylated liposomes (BLP) with encapsulated DNA or Quantum Dots, and the physicochemical characterization of the particles according to size, biotinylation and DNA or QDs encapsulation efficiency.
The second section presents the EGF-based targeting strategy of the BLP either loaded with DNA or QDs and then focuses on the cell binding experiments. A two-color BLP labeling strategy usign QDs was designed, where colocalization of green emitting QD525 and red emittingQD655, allowed the immediate recognition of targeted BLP during live cell imaging by confocal microscopy.
The specificity for the EGF receptor was examined in a competitive binding assay in the presence of excess ligand EGF and after incubation of BLP with cell lines devoid of the receptor.
Transfection competencies of EGFR-targeted and DNA loaded BLP particles were evaluated as the percentage of cells expressing GFP compared to cells incubated with non- targeted particles.
The third section presents the non-covalent DNA labeling strategy using bis-PNA and a single QD per plasmid and describes the characterization of the labeled complexes by gel electrophoresis and Atomic Force Microscopy. The data, although preliminary, are complementary to the main results of this thesis and represent the first steps towards a QD- labeled plasmid that could be monitored along the pathway to the nucleus. Such a labeled plasmid appears as an attractive cargo for the BLP.
2. Materials and methods
2.1. Reagents
2.1.1. Lipids
Lipids were purchased from Avanti Polar Lipids (Alabaster, USA) and from Northern Lipids (Vancouver, Canada) as lyophilized powders or dissolved in chloroform. Chemical structure and nomenclature are depicted in Table 2.1. Lipids were stored under argon in a desiccator at -20 °C.
Description Nomenclature Chemical structure
Phospholipid
DOPE
1,2-Dioleoyl-sn-Glycero-3- Phosphoethanolamine
Cationic lipid
DOTAP 1,2-Dioleoyloxy-3- trimethylammoniumpropane
chloride
Biotinylated lipid
Biotin-PEG-DSPE 1,2-Distearoyl-sn-Glycero-3-
Phosphoethanolamine-N- [Biotinyl (Polyethylene Glycol) 2000] (Ammonium
Salt)
PEG- modified ceramide
PEG-Cer-C20 N-Arachidoyl-Sphingosine-1-
[Succinyl (Methoxy (Polyethylene Glycol) 2000)]
PEG- modified ceramide
PEG-Cer-C8 N-Octanoyl-Sphingosine-1-
[Succinyl (Methoxy (Polyethylene Glycol) 750)]
pH-sensitive lipid
PEG-DSGS 3-O-[2’-(omega- monomethoxy (Polyethylene
Glycol) 2000) succinoyl]-1,2- Distearoyl-sn-Glycerol
Table 2.1: Synthetic lipids used in the formulation of lipid particles.
Chloroform, Ethanol, Methanol, Isopropanol and Water HPLC quality were purchased from Merck.
Buffers were prepared with analytical grade chemicals. Detergent, Triton X-100 was purchased from Fluka, detergent 1-o-n-octyl-β-D-glucopyranoside (OGP), sucrose, HABA 4- Hydroxyazobenzene-2-carboxylic acid) and kanamycin from Sigma and nonpolar polystyrene adsorbents Bio-Beads SM-2 from BioRad.
UltraPure™ Agarose was purchased from Invitrogen.
Luria Bertani (LB) medium in powder form was obtained from Difco.
2.1.2. Biotin and biotin-related reagents
Biotin and Biocytin were purchased from Sigma, Ez link Sulfo NHS-LC-LC-Biotin and EZ-Link PEO-Biotin Dimer (+)-Biotinyl-hexaethyleneglycol dimmer from Pierce.
Biotin-EGF and EGF (human recombinant) were obtained from Molecular Probes. 2.1.3. Buffers
Hepes buffered saline (HBS): 10 mM HEPES (2-hydroxyethyl)-1-piperazine ethanesulphonic acid, 150 mM NaCl (sodium chloride), pH 7.4.
Hepes-Mg buffer: 10 mM HEPES (2-hydroxyethyl)-1-piperazine ethanesulphonic acid, 5 mM MgAc2 (Magnesium Acetate), 10 mM KCl pH 7.4 prepared with Ultrapure Water HPLC quality and filtered through 0.02 µm pore size.
Phosphate buffered saline (PBS): 137 mM NaCl, 2.7 mM KCl, 7.9 mM Na2HPO4,1.5 mM KH2PO4, pH 7.3.
Tyrode’s buffer: 135 mM NaCl, 10 mM KCl, 0.4 mM MgCl2, 1 mM, CaCl2, 10 mM HEPES, pH 7.2.
20 mM Glucose and 0.1% BSA were added before use.
10 mM Sodium Acetate buffer pH 5.2: 1g/l Sodium acetate tri-hydrate and 0.16 g/l acetic acid (glacial).
50X TAE: 242 g Tris base, 57.1 g glacial acetic acid, 100 ml of 0.5 M EDTA, pH 8.0.
2.1.4. Quantum Dots
Streptavidin-coated QDs (StAv-QDs) with maximum fluorescence emission peaks at 525 nm (StAv-QD525), 605 nm (StAv-QD605) and 655 nm (StAv-QD655) and ITK-carboxyl QD655 were purchased from Invitrogen.
2.1.5. Labeled proteins and organic fluorophores
Streptavidin Alexa Fluor488 (ex. 488 nm/em. 519 nm), Transferrin Alexa Fluor488 , Sybr Green (ex. 495 nm/em. 521 nm), Cy5 NHS ester (ex. 650 nm/ em. 670 nm) and Hoechst 33342 (exc.
365 nm/ em. 455 nm) were purchased from Molecular Probes.
2.1.6. Plasmid DNA
Plasmid pEGFP-C1, 4.7 Kb encoding the Enhanced Green Fluorescent Protein (EGFP: ex.
488 nm/em. 507 nm), was bought from Clontech Laboratories (Heidelberg, Germany). This encodes a red-shifted variant of the wild-type green fluorescence protein.
2.1.7. bis-PNA (bis Peptide Nucleic Acid)
bis-PNA (bis- Peptide Nucleic Acid) MW: 5655 (ε260 nm= 107840 M-1cm-1) was acquired from Oswel, Southampton, UK.
bis-PNA sequence: NH2-LL-TCCCCCTTT-LLL-TTTCCCCCT-LL, where L represents 8-amino-3,6- dioxaoctanoic acid.
bis-PNA anchoring sequence in p-EGFP C1 plasmid: AGGGGGAAA at base pair position 4515-4523.
2.1.8. Restriction endonucleases and DNA standards
Restriction endonucleases Apa LI, Ava II and Ase I and the corresponding buffers were obtained from New England Biolabs. DNA markers of 100 bp and 1 Kb were from Invitrogen.
2.2. Cell lines, media and labware
Human epidermoid carcinoma cell line A431 and CHO (Chinese hamster ovary) cell line were obtained from the American Tissue Culture Collection, while the melanoma cell line WM983A, was kindly provided by Prof. Meenhard Herlyn (Wistar Institute, Philadelphia, PA).
Dulbecco’s Minimal Essential Medium was obtained from Gibco. Leibovitz’s L15 medium was from Mediatech, Inc. FBS was bought from PAN. Penicillin and Streptomycin and Trypsin were purchased from Gibco. EDTA, BSA and glucose were from Sigma.
Paraformaldehyde (PFA) of analytical grade was freshly prepared as a 4% solution in PBS for cell fixation.
Glass coverslips 12 mm were acid washed and sterilized. LabTek 8-well or 2-well glass chambers, plastic cell culture dishes and flasks were purchased from Nunc or Sarstedt.
2.3. Miscellaneous
Ultraconcentrators 3 kDa and 10 kDa cut-off, Amicon (Millipore), Vivaspin (Viva Science Sartorius, Göttingen)
Gel filtration columns Sephacryl-HR 200 and HR-400 and Superdex™ 75 /Superdex™ 200 (GE Healthcare)
Native 3-8 % Polyacrylamide precasted gels (Invitrogen)
Dialysis Cassettes, Slide-A-Lyzer, 10 kDa cut-off for 0.5-3 ml sample, (Pierce)
Streptavidin-coated magnetic beads Dynabeads M-280, (Dynal Biotech)
Thin-wall polypropylene centrifuge tubes 2.5 ml (Beckman)
Filters set for fluorescence microscopy (Chroma and Omega)
Silicon Nitride ultrasharp cantilevers for AFM (Veeco and Mikromarsch)
2.4. Equipments and softwares
Bench top centrifuge Mini spin plus (Eppendorf)
Centrifuge Beckman Avanti J-25
Ultracentrifuge/rotor with swinging
buckets TL-100/TLS 55 Beckman
Chromatographic system SMART, Pharmacia Biotech, GE Healthcare Life Sciences
Laser-based scanning system for
fluorescence detection Typhoon, Amersham Biosciences UV-Vis spectrophotometer Cary 100 Scan
Fluorescence spectrophotometer Varian Cary Eclipse
Particle analyzer Zeta Sizer Nano ( Malvern Instruments) Atomic Force Microscope Nanoscope IIIa Multimode AFM (Veeco)
Electron microscope Philips CM 12
Optical microscopes Axiovert 100 (Zeiss, Göttingen, Germany) Olympus IX71
iXon CCD camera for microscopy Andor technology
Laser Scan Confocal Microscope LSM 510 Meta (Zeiss, Göttingen, Germany)
Flow cytometer Coulter Epics
Software for Image analysis and data handling:
AFM analysis: Vision 700 Image analysis:
Image J free software for microscopy DIPImage toolbox in Matlab 7.01
Flow cytometry analysis: Reflex software
Nanoscope v. 7.00. (Veeco)
(http://rsb.info.nih.gov)
(http://www.ph.tn.tudelft.nl/DIPlib/index.html) (http:// www.freewebs.com/cytoflex)
2.5. Preparation of Biotinylated Lipid Particles (BLP)
As depicted in the diagram below (Figure 2.1), to prepare 1 ml of BLP formulation, 8.4 µmols DOPE (neutral lipid), 0.6 µmols DOTAP (cationic lipid) and 0.024 µmols Biotin-DSPE were initially mixed in 2 ml glass vials. In individual vials containing this lipid mixture, different micromolar amounts of PEG-lipids (Table 2.1) were added to obtain 1.4, 2.7 and 10 (mol%) PEG content and a final lipid concentration of ~10 mg/ml in the BLP formulation. These solutions were dried with a stream of argon and residual chloroform was removed under vacuum overnight.
Figure 2.1: Flow diagram of BLP preparation with encapsulated plasmid DNA or QDs.
2.5.1. Detergent dialysis technique
The resulting lipid film was hydrated in 0.5 ml of HBS containing 100 µl of 1M OGP detergent with continuous magnetic stirring at 60 °C to favour complete dissolution. For the preparation of BLP with encapsulated DNA (BLP-DNA), the detergent dialysis technique described by (Hofland et al., 1996) was applied with modifications. In this approach, 100 µg or 200 µg peGFP-C1 plasmid (purified as described in 2.10.1) were diluted in 0.5 ml HBS and added with stirring to the solubilized lipids-detergent solution (0.5 ml). The mixture was keptat
RT for 30 min and followed by a dialysis at RT against HBS using a 10 kDa MWCO membrane for 2-3 h or until turbidity was noticeable. Then, the dialysis was continued in a cold room (~4
°C) overnight and after two buffer changes of 1 l HBS, residual detergent OGP was removed by adding polyestyrene SM2 Biobeads (adsorbent capacity ~117 mg OGP/g beads). Control BLP samples without plasmid were prepared in parallel.
In order to prepare BLP with encapsulated QDs (BLP-QDs), ITK-carboxyl QD655 were added to the corresponding lipid-detergent mixture in HBS (50 nM final QDs concentration). Samples were immediately dialysed as described for BLP-DNA particles.
2.5.2. Ultracentrifugation in sucrose density gradient
In order to separate and recover DNA and QDs loaded BLP from less dense empty BLP and non encapsulated material, the samples obtained after the detergent dialysis were loaded on top of a discontinuous sucrose density gradient.
The sucrose gradient was prepared in 2.5 ml Beckman ultraclear thin-wall centrifuge tubes by carefully applying with a tip or syringe, layers of 0.6 ml 2.5 %, 10%, and 20% (w/v) sucrose in HBS (Mok et al., 1999). The BLP samples recovered from the detergent dialysis were adjusted to 1% (w/v) sucrose density and were added on top of the gradient. Ultracentrifugation was carried out for 5 h at 160,000 g at 10 °C in a TLS-55 rotor with swinging buckets.
The resolved turbid bands from BLP-DNA samples were carefully recovered and, together with aliquots of the gradient, were analyzed for particle size, biotin content and DNA encapsulation efficiency.
Centrifuge tubes containing BLP-QDs samples were briefly illuminated with UV light to detect QD655 fluorescence, and consequently the location of QDs containing bands.
Fluorescent bands were analyzed by TEM at the Electron Microscopy Department of the institute.
2.5.3. Particle size analysis by Dynamic Light Scattering
The mean hydrodynamic diameter of the obtained BLP was measured by Dynamic light scattering (DLS) which involves the determination of how the intensity of the light scattered by a solution of moving particles varies with time. This variation is correlated with the speed at which particles move, which can be characterized by their diffusion coefficients (Pecora and Aragon, 1974). The hydrodynamic diameter d of particles is obtained from the diffusion coefficients D, according to:
d = kT
3π η D
where, k is the Boltzman constant, T is the temperature and η is the solvent viscosity.
DLS measurements were performed with a NanoZetasizer from Malvern Instruments,
Samples were diluted in PBS and the measurements performed at RT. The PBS used to dilute the samples was previously filtered through 0.02 µm pore size to eliminate potential interfering impurities.
The operating protocol was chosen for spherical particles, applying 12 to15 runs per measurement, which were automatically selected according to the concentration of particles in the sample, and three measurements were performed on each sample. The autocorrelation function for size distribution was calculated using the CONTIN mathematical approach for heterodisperse, polydisperse and multimodal systems (Provencher, 1982).
The mean hydrodynamic diameter obtained represents only an intensity-based average value and does not give any information on the prevailing size distribution. For this reason, the polydispersity index (pdi) is also stated to give information about the actual distortion of a monomodal distribution. The pdi can have values between 0-1 and is equivalent to the variance σ2 of the size distribution. Samples with pdi <0.25 are considered as monodisperse solutions.
2.5.4. DNA quantitation
The amount of plasmid recovered and encapsulated after dialysis and ultracentrifugation in sucrose density gradient was quantitated using the Sybr Green fluorescence assay (Zhang et al., 1999). In the presence of DNA, the dye binds specifically to double-stranded DNA and emits at 522 nm. Briefly, 25 μl aliquots of standard solutions varying from 0-5 µg/ml plasmid DNA or dilutions of BLP were mixed with 25 μl of HBS buffer containing 0.5 μg/μl Sybr Green solution. Sybr Green fluorescence was excited at 495 nm and emission spectra were collected from 505 to 700 nm with a slits width of 10 nm for both excitation and emission.
The plasmid DNA content in BLP formulations was calculated from the linear curve fitting obtained after plotting the maximum fluorescence intensity (a.u.) of Sybr Green at 522 nm as a function of DNA concentration in µg/ml plasmid (Figure 2.2).
Figure 2.2: Representative DNA calibration curve using Sybr Green dye. Sybr Green excited at 495 nm and emission recorded from 505-700 nm. Maximum fluorescence intensity (a.u.) at 522 nm was plotted as a function of plasmid DNA concentration. Data correspond to three curves obtained in
independent experiments. Each point is an average of duplicates. Free Sybr Green is non-fluorescent in HBS (0 µg/ml plasmid DNA). The presence of 0.1% Triton-X100 did not change the maximum fluorescence intensity at 522 nm of Sybr Green upon DNA binding (not shown).
Free or exposed plasmid DNA was quantitated directly in BLP solutions. Then, lipid particles were solubilized with 0.1% Triton X-100 and total DNA content was measured. The obtained value was related to the initial amount added to compute DNA encapsulation efficiency.
2.5.5. Biotin quantitation
The biotin content of BLP was determined on samples recovered after detergent dialysis and ultracentrifugation in sucrose gradient.
The fluorometric assay used is based on Foster resonance energy transfer (FRET) (Batchelor et al., 2007), providing high sensitivity to detect nanomolar concentrations of biotin linked to proteins or nucleic acids. In this assay (Figure 2.3A), the Alexa Fluor488-streptavidin conjugate (StAv-Alexa488) acts as FRET donor and the dye HABA as a quencher, occupying the biotin binding sites of the dye-labeled streptavidin. In the absence of biotin, HABA quenches the fluorescence emission of the Alexa Fluor488 dye via FRET. When biotin or a biotinylated molecule is added, HABA is displaced from the biotin binding sites resulting in an increase in the donor fluorescence intensity proportional to the amount of biotin present in the sample.
The standard curve of biocytin, a water-soluble biotin analogue, was obtained by titrating a solution containing 25 nM StAv-Alexa488 and 125 μM HABA in HBS. Biocytin was added from an aqueous stock solution to give final concentrations of 0–200 nM in 50 μl samples.
Fluorescence of standards and samples was measured in a spectrofluorometer, Alexa Fluor488
dye was excited at 485 nm and emission spectra were collected between 500-700 nm.
The fluorescence intensity at the emission maximum (519) nm was plotted as a function of biocytin concentration (nM) and the data were fitted to a sigmoid curve to obtain the corresponding values for the unknown samples (Figure 2.3B). In the conditions assayed a ~6- fold increase in fluorescence signal upon complete displacement of HABA was observed with a detection limit of approximately 3 nM biocytin (Figure 2.3C).
A
biotin Streptavidin
Alexa Fluor 488
HABA
B C
Figure 2.3: Fluorometric assay for biotin quantitation. A: General scheme of the assay adapted from (Batchelor et al., 2007). HABA acts as acceptor of Alexa Fluor488 fluorescence while bound to biotin binding sites. When biotin or biocytin is added, HABA is displaced and the fluorescence signal of Alexa Fluor488 increases. B: Biocytin standard curve obtained by titrating a solution containing 25 nM StAv- Alexa488 and 125 μM HABA in HBS. At least 8 values were obtained to plot the data within a biocytin range of 0-200 nM. Excitation of Alexa488 at 485 nm and emission collected from 500 to 700 nm. The data illustrate a ~6-fold increase in fluorescence signal upon complete displacement of HABA.
According to manufacturers, three biotin-binding sites are available per streptavidin molecule corresponding to 75 nM biotin binding sites C: The detection limit achieved was approximately 3 nM biocytin.
2.5.6. Transmission Electron Microscopy
The size and morphology of BLP and encapsulated QDs were analyzed by TEM by Dr.
Dietmar Riedel in the Electron Microscopy department at the MPIbpc.
2.6. Labeling and targeting of BLP
2.6.1. EGF-QDs preformed complexes
The formulation of the streptavidin conjugated QDs includes a polyethylene glycol 2000 (PEG) outer layer to which 6-8 streptavidin molecules per QDs are covalently linked. This allows variations in the stoichiometry of the ligand-QD complexes. Procedures for binding biotinylated ligands to commercial streptavidin conjugated quantum dots were optimized in our lab and detailed protocols are described in (Lidke et al., 2007).
Biotin-EGF were coupled to StAv-QD525 (or QD605) at ratio 2:1 or 4:1 in HBS with 0.2% BSA at 50 nM QDs final concentration. Solutions were incubated at 15 °C for at least 30 min with gentle agitation before coupling to BLP.
2.6.2. EGF-QDs coupling to BLP
BLP loaded with QD655 or plasmid DNA were incubated with preformed complexes of EGF- QD525 (1:1 v/v) in HBS suplemented with 0.2% BSA. Final concentration of QD525 was 25 nM whereas final biotin concentrations on BLP varied from 0.2-10 µM biotin. This mixture was incubated for at least 2 h at 15 °C with continuous shaking.
Separation of EGF-QDs-BLP from EGF-QDs could not be achieved by size exclusion chromatography or sucrose gradient (data not shown) and resulted in considerable sample loss. Therefore, upon incubation with cells, residual free preformed complexes (EGF-QDs) were expected to bind to EGF receptors (Lidke et al., 2004) being indistinguishable from the targeted particles (EGF-QD–BLP). In order to solve this problem, two-color BLP labeling strategies were designed for BLP either loaded with QDs (Figure 2.4A) or plasmid DNA (Figure 2.4B).
Figure 2.4: EGF tagged and two-color labeled BLP. A: with encapsulated carboxyl-QD655. B: with encapsulated plasmid DNA.
In the case of BLP-DNA (Figure 2.4B), particles were incubated first with EGF-QD525 in HBS supplemented with 0.2% BSA for 30 min at 10 °C. Then, the non-targeted StAv-QD655 were added to the EGF-QD -BLP particles, and incubation proceeded for at least 2 h at 10 °C
before in vitro cells experiments. The final concentration of both StAv-QDs was 25 nM whereas the final concentration of encapsulated DNA varied among preparations and is stated in each experiment.
In both approaches, the green emitting StAv-QD525 are the ones targeted to the EGFR in A431 cells, whereas the red emitting QD655 without the ligand, serve to track the BLP themselves.
2.7. Live-cell experiments
2.7.1. Cell culture
A431 and CHO cell lines were cultured in Dulbecco’s Minimal Essential Medium (D-MEM) supplemented with 10% fetal calf serum (FCS), 50 U/ml of penicillin and 50 U/ml of streptomycin. WM983A cells were grown in three parts DMEM and one part Leibovitz’s L15 medium supplemented with 10% heat inactivated (56 °C for 20 min) FCS and Pen-Strep. All cells were grown as a monolayer at 37 °C in a humidified atmosphere containing 5% CO2. For microscopy experiments, cells were seeded one or two days in advance onto 12-mm- diameter glass coverslips and employed at a 40-50% confluence. For binding experiments, cells were starved in serum free DMEM for 4 h prior to the incubation with BLP to reduce signaling induced by growth factors present in the serum.
2.7.2. Experimental conditions for the incubation of A431 cells with BLP
Two-color QDs labeled, EGFR-targeted and non-targeted BLP-DNA or BLP-QDs were diluted in 200 or 500 µl Tyrode buffer with 0.5% BSA, to obtain 0.5 nM QD525 and 2 nM EGF final concentration. Preincubation of targeted and non-targeted BLP with starved A431 cells was carried out first at 15 °C, a temperature non-permissive for endocytosis. This preincubation time was varied from 5 to 30 min, then cells were warmed at 37 °C and were incubated from 10 min to 2 h. Finally, unbound BLP were removed and cells washed with Tyrode’s buffer.
Confocal fluorescence microscopy was performed on live cells immediately after removing the complexes and after 30 min and 1, 2, 4, 12, 24 and 48 h incubation at 37 °C.
When incubation was extended for more than 1h after removing the BLP, Tyrode’s buffer was replaced by complete DMEM.
2.7.3. Incubation of EGF-QDs and Transferrin-Alexa488 complexes with A431 cells
A431 cells were grown on glass coverslips at a density 40-50 % and serum-starved before experiments. Transferrin-Alexa Fluor488 (60 nM) and EGF-QD605 complexes (2 nM) were added to the cells to allow simultaneous internalization. Incubation with the complexes at 37 °C proceeded for 10 and 30 min.
In a second experiment, EGF-QDs complexes were added first to the cells and after 2 h incubation at 37 °C, Transferrin-Alexa Fluor complexes were added for 5 min. Then,
